OpenVLM MMT-Bench - Order Hallucination: leaderboard
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-8B | 65 |
| 2 | Claude 3 Haiku | 55 |
| 3 | Llama 3.2 90B Vision Instruct | 55 |
| 4 | Pixtral-12B | 50 |
| 5 | Grok 2 (1212) | 50 |
| 6 | InternVL3-78B | 45 |
| 7 | InternVL2-8B | 45 |
| 8 | Qwen 2 VL 72B | 45 |
| 9 | InternVL3-38B | 45 |
| 10 | Qwen 2 VL 2B | 45 |
| 11 | GPT-4.1 Nano | 45 |
| 12 | Gemma 3 12B | 45 |
| 13 | GPT-4.1 (2025-04-14) | 40 |
| 14 | GPT-4.1 Mini | 35 |
| 15 | Qwen 2 VL 7B | 35 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-order-hallucination · How It Works · Data refreshed daily, snapshot 2026-09-05.